RVC Text to Speech:
any voice reads your text
Type a line, pick any RVC model in your Library, and Xport Studio speaks it in that voice. No Python, no command line, nothing uploaded. It runs on your own Mac or PC.
macOS 15+ (Apple Silicon) · Windows 10+. Text to speech with RVC voices is free for 7 days, then part of Xport Pro, $79 one time. Never a subscription.
Hear it: one line, five voices
The same sentence, read by Xport's text to speech and converted by different RVC models. Tap a name to hear it.
Your RVC voices can read text. Type any line, pick any RVC model in your Library, and Xport speaks it in that voice.
Watch a line get read.
This is the Text to Speech screen in Xport Studio. Type a line, pick a voice from your Library, RVC models included, and press Generate.
How RVC text to speech works
RVC is a voice converter, not a text to speech engine: it needs audio to start from. Xport Studio chains the two steps for you, so one click turns text into speech in your RVC voice.
1. A neural voice reads your text
Xport's built-in text to speech speaks the line, picking a male or female base voice to match your model.
2. Your RVC model converts it
The speech runs through your .pth model (and its .index file), so it comes out in that voice.
3. You get a WAV in that voice
Play it back, A/B it, or drag the file straight into your DAW or video editor.
All on your computer
No upload, no queue, no credits. After the one-time downloads it works offline.
How to make an RVC model read text in Xport
- Add your RVC voice. Import a .pth model (plus its .index file if you have one) into your Library, or train your own voice with voice cloning.
- Type your line. Open Text to Speech and type or paste anything: a hook, an intro, a script, a whole voiceover.
- Pick the voice and generate. Choose the RVC model, press Generate, and export the WAV.
The easy way to use RVC with text to speech
You can wire RVC to a text to speech engine yourself: a Python project, a WebUI, or an extension inside another app. It works, but you install Python, download models by hand, and pass audio from one tool to the next. Xport does the whole chain in one window.
Do it yourself
Install Python and dependencies, find a text to speech model, point a script at your .pth, convert, then clean up the files.
Xport Studio
Install one app, import your model, type, press Generate. Same models, no setup, on Mac or Windows.
What people make with it
Voiceovers
Narrate videos and shorts in a custom voice without recording a single take.
Your own voice, typed
Clone your voice once, then type lines instead of re-recording every fix.
Intros and drops
Spoken intros, tags and ad-libs for tracks and mixes, in the voice you want.
Character dialogue
Prototype lines for games, animations and mods before casting real actors.
Podcast segments
Read-outs, sponsor slots and recaps in a consistent host voice.
Song ideas
Hear a lyric spoken in an artist voice to test the flow before anyone sings it.
FAQ
Can RVC do text to speech on its own?
No. RVC (Retrieval-based Voice Conversion) changes one voice into another, so it needs audio to start from. Xport Studio adds the missing step: a neural text to speech voice reads your line first, then your RVC model converts it, all in one click.
Which RVC models work?
RVC voice models saved as .pth files, with or without the matching .index file. Import them into your Library once and they show up in Text to Speech and in the vocal converter. Voices you train yourself in Xport work too.
Do I need Python or a command line?
No. There is nothing to install besides the app: no Python, no WebUI, no model paths to wire up. Import the model, type your text and press Generate.
Is RVC text to speech free?
Text to speech with RVC voices is one of the AI tools, which are free for 7 days. After that it is part of Xport Pro, $79 one time, never a subscription. The stem splitter, key and BPM finder, converter and the other tools stay free forever.
Does it work offline?
Yes. After the one-time model downloads, text to speech and the RVC conversion both run on your computer with no internet connection.
Is my text or audio uploaded?
No. Your text, the generated speech and your voice models stay on your machine. Nothing is sent to a server and nothing is used to train AI.
How fast is it?
A short line takes a couple of seconds. On an Apple Silicon Mac the RVC step for a five second line runs in about a second and a half.
Can I use any voice I find online?
Only use voices you have the right to use, such as your own voice, a voice you were given permission for, or a model licensed for your use.